Killed by design: Consequences of Poor UX & How to Do Better

With great impact come great failures too. Read on for three examples of UX design gone horribly wrong and learn how not to kill your users with bad design.

Yes, the title of this article has drama written all over it. I mean, everybody likes drama, right? And no, the following lines are not about the near-term future where terminators terminate and some people even teleport naked. In fact, the examples we will be describing here might come off as a bit more surreal.

All set, right? Or maybe just that one other thing you might be thinking, which is: “What the hell is UX anyway?” Well, I’m glad you asked:

User Experience design represents the quality of an experience you have when interacting with a certain product, be it physical or digital. When done well, it becomes invisible to your users. It also becomes delightful and - most importantly - helpful. Conversely, when executed poorly, it sticks out like a fox in the hen house. Not good for your users, definitely not good for hens. Especially if you own the hens. In the best case scenario, bad design will “only” frustrate and anger your users - no big deal, right? But what happens when it becomes so bad that it can virtually kill?

How Bad UX Killed Jenny

This particular example became well known in the design world thanks to a book called “Tragic design”, by Jonathan Shariat & Cynthia Savard Sauicer, in which they highlight different instances of how bad design can harm. Great read, by the way (find it here).

Since the patient’s name was never actually shared, we will call her Jenny (following the authors’ example). After battling cancer for four years and finally being discharged, Jenny was used to spending a lot of time in the hospital. Unfortunately, some time after she relapsed and had to be given a very strong chemo treatment medicine. This medicine is so strong and so toxic that it requires pre-hydration and post-hydration for three days with a special kind of fluid.

Furthermore, everytime the medicine was administered, three nurses had to attend to the charting software to enter all the requirements and make the appropriate orders - and this is where they missed a very critical piece of information. In short, Jenny was supposed to be given 3 days of post-hydration fluid, but the nurses, with over 10 years experience, were so distracted trying to figure out how to use the software that they completely missed this information.

When the morning nurse came in the next day, Jenny had died of toxicity and dehydration. For two shifts, she had missed her hydration treatment, all while the three experienced nurses were too busy trying to figure the software out.

How Bad UX Killed Anton Yelchin

It was in 2016 that the death of Anton Yelchin shook the world. The 27-year-old was successful Hollywood actor known - among other roles - for playing Commander Chekov in the Star Trek movies.

He was found crushed between his car, a 2015 Jeep Grand Cherokee, and the gate at the end of the driveway. What happened? It appears that he had exited the car and walked behind it to close the gate, believing the transmission was in “Park”. Instead, it was actually in “Reverse” or “Neutral”, and the car rolled down his steep driveway, killing him.

The National Highway Traffic Safety Administration investigated Fiat Chrysler vehicles in February after drivers complained about vehicles rolling away while in park. Apparently, the transmission shifters in these cars function differently than standard shifters. Instead of a gear lever that physically moves from the "drive" position to the "park" position, these levers remain in the same place as letters like "D" and "P" light up. This caused many drivers to leave their vehicles thinking they'd put their cars in park when they hadn't successfully done so—resulting in 41 injuries, 212 crashes and 308 total incident reports.

Due to this bizarre accident and an obvious design flaw, Fiat Chrysler Automobiles decided to recall 1.1 million SUVs in response.

How Bad UX Killed 87 People

On the cold night of January 20, 1992, Air Inter Flight 148, commanded by Captain Christian Hecquet and First Officer Joël Cherubin, departed Lyon–Saint Exupéry Airport in Lyon, France. Both were experienced pilots with over 12,000 hours of flying between them.

Due to severe weather conditions, the air traffic controller suggested they head to an alignment beacon. The captain agreed and started to calculate the new descent approach for landing, programming the autopilot in the process . He properly calculated a smooth descent angle of 3.3 degrees, which he entered into the instrument. He then made the final turn to align with the runway and beacon, corrected his direction, and initiated the landing sequence he had programmed in.

Everything was going according to plan, right until the cloud cover broke and they came face to face with the mountain. Within a few seconds, the aircraft struck trees and impacted into an 826-meter-high (2,710 feet) ridge near Mt. Sainte-Odile.

87 people were killed on impact while 9 of them - miraculously - survived, sustaining only injuries.

What caused all this? As mentioned, due to weather conditions, pilots were unable to rely on technology to initiate a descent, so calculations had to be done manually. So far so good. However, due to poorly executed interface design, pilots confused the types of input needed initiate a proper descent.

The interface had two modes, and although both are for descend, they required different units. The only difference between the two is the presence of a decimal and small letters displayed just above them. The pilot did not differentiate properly between the two, even after setting up the parameters, due to the mentioned similarity between numbers. It took just one minute after that to crash the plane straight into the mountain.

How To Do Better

Besides the fact that all of these examples end tragically, the single most important thread that ties them together (as in many, many other cases) is this: an impressive lack of consideration for those who will actually use the product.

No matter if we are talking about nurses handling the life-saving software, or pilots operating a multi-million equipment with seemingly same amount of buttons and switches in the cockpit, or even just the regular folks - like you and me - that want to have the basic commodity of driving their car and feel safe doing so, it is worth remembering the following notion:

User needs come first, technology second.

Sounds simple, but what does that actually mean? Again, I’m glad you asked, because by asking questions, such as you are doing now, dear reader, is where all true design efforts should come from. In particular, you need to:

Doctor: 'What are you doing here, honey? You're not even old enough to know how bad life gets.'
Cecilia: 'Obviously, Doctor, you've never been a thirteen-year-old girl.'— The Virgin Suicides

1. Go out and listen to those you are designing for

And no, this doesn’t mean that you should ask people what they like. I like samurai swords, that doesn’t mean I’m gonna buy one online. At least not until I buy a nice kimono first and grow hair like Tom Cruise in that samurai movie.

By conducting user research, we want to see how people really use the product. It is a process where they show you what they do, not tell. Yes, people are confusing.

2. Test your designs, and then test them again

Always try to test your designs, concepts and whatever you have, whenever you can. Ideally, these should be done with your target audience, but even asking your work colleagues to spare a few minutes for a quick test is better than not doing any. Also, this is your unique chance to put on a lab coat you stole in escape room that one time and show your colleagues how truly important your work is.

3. Put things into the right perspective

Ask yourself: how is this design going to affect people? How is it going to affect the environment we all live in? Are the benefits of the product I am building outweighing the potential negatives? This also means that we have to constantly ask questions about what we as designers do, and to take responsibility for what we put out in the world. There will be misfires, sure, and because of that it is all the more important to embrace such mistakes and learn from them.

By taking all these factors into account, you will have all the right ingredients to create something meaningful - something designed with people, for the people. “Panic bars” on exit doors are just one example of things done right when the design process begins and starts with your users.

People fleeing a fire would die if they encountered exit doors that opened inward, because they would keep trying to push them outward, and when that failed, they would push harder. Great strategy is to design the doors so that they open when pushed.

Also, it is important to mention that following all the steps above doesn’t mean success of your product is 100% guaranteed, but it does mean that the chances of it becoming a failure are dramatically reduced.

As a final take away, just remember this: the experiences we create as UX designers affect real people in real situations. Being the gatekeepers of technology, it is up to us to ensure the gates are as wide open and accessible as possible. To be good designers, we cannot simply design for success alone, we must always design for failure; after all, it is not a matter of if, it is a matter of when.

TOMISLAV KOLAR / August 12, 2022